Method
Survey of 12 October 2026, made on each project’s public page: the GitHub repositories of Pulse, PegaProx and ProxCenter, and the ProxCenter website. Every cell links to its source. When a feature does not appear on those pages, the table says “not described”, not “missing”: the tool may offer it elsewhere or have added it since. We neither installed nor measured these tools for this comparison, and we publish one of them: ProxTrace. Names are cited for identification; corrections are welcome at contact@proxtrace.com.
Criterion by criterion
- Yes
- Partial : limited, or only in a paid edition
- No
- Not described : not on the public page we read, which does not prove the tool lacks it
| Criterion | ProxTrace | Pulse | PegaProx | ProxCenter |
|---|---|---|---|---|
| Configuration change history, before and after values | The core of the product, with the raw trace | Not described on its public page (source) | Drift: nightly snapshot compared with a baseline (source) | Not described on its public page (source) |
| Who made a change: account and source address | Matched only when unique (“Correlated”) | Not described on its public page (source) | Not described on its public page (source) | Not described on its public page (source) |
| Audit log | Proxmox VE tasks, API calls and logins, actions in the console | In the Pro edition (source) | HMAC-signed (source) | Not described on its public page (source) |
| Proxmox best-practice checks | 115 checks, four statuses including “Not assessable” | Pulse Patrol: backups, capacity, restarts, clocks (source) | CIS audit and hardening score (source) | Not described on its public page (source) |
| Mapping to NIS2, ISO 27001, BSI or SOC 2 | Evidence provided, no ready-made mapping | Not described on its public page (source) | Read-only compliance dashboard, PDF export (source) | Not described on its public page (source) |
| Sealed or exportable reports | Sealed reports, export from the console | Reporting in the Pro edition (source) | Compliance PDF export (source) | Advanced reporting in Enterprise (source) |
| Vulnerability (CVE) scanning | Version and repository tracking only | Not described on its public page (source) | Per node, with debsecan (source) | In Enterprise, according to the FAQ (source) |
| Real-time monitoring | Not a monitoring tool; checks re-evaluated every five minutes | Its core purpose (source) | Live metrics dashboard (source) | From the free edition (source) |
| Alerts and notifications | Notifications for changes that matter | Alerts, including ntfy, Gotify and Pushover (source) | Alert on configuration drift (source) | Email in Community, multi-channel in Enterprise (source) |
| Read-only: no action possible on the infrastructure | Read-only agent, changes nothing | Read-only token; governed fixes in Pro (source) | Management tool: acts on machines (source) | Management tool: consoles, migrations (source) |
| Management actions (start, migrate…) | None, by design: administration stays in Proxmox VE | Policy-governed fixes, in Pro (source) | Start, stop, migration, one-click hardening (source) | Consoles, cross-hypervisor migration, DRS in Enterprise (source) |
| Several clusters or clients | Separate organisations, ADMIN, OPERATOR, AUDITOR roles | MSP edition, several client workspaces (source) | Unified dashboard (source) | Unlimited clusters in Community (source) |
| AI assistant | Optional, on demand, cites its sources | Patrol, assistant and MCP adapter (source) | Not described on its public page (source) | Analytics and predictions in Enterprise (source) |
| Open source | Not at this time | MIT; Pro edition is commercial (source) | AGPL-3.0 (source) | Community free, buildable from source; EULA in the repository (source) |
| Usable for free today | Not yet commercially available | Community edition (source) | Free, funded by donations and sponsors (source) | Community edition (source) |
| Self-hosting | Self-hosted offer presented, in preparation | Docker, Compose, Helm, LXC (source) | Script, APT repository, Docker (source) | Docker with PostgreSQL (source) |
| Platforms other than Proxmox | Proxmox VE only | Docker, Kubernetes, TrueNAS (source) | XCP-ng (source) | Migration from ESXi, XCP-ng, Nutanix, Hyper-V (source) |
What the others do better
- Real-time monitoring: it is Pulse’s job, and ProxCenter does it in its free edition. ProxTrace does not monitor metrics and is not built to alert within the minute on capacity pressure.
- Management: PegaProx and ProxCenter start, stop and migrate machines from their console. ProxTrace takes no action, by design; administration stays in Proxmox VE.
- Compliance mapping: PegaProx maps BSI, ISO 27001, NIS2 and SOC 2, with a CIS audit. ProxTrace assesses Proxmox VE best practices and produces evidence, without a ready-made mapping to those frameworks.
- Being available now, as open source: Pulse (MIT) and PegaProx (AGPL-3.0) can be installed today for free, and Pulse has the largest community of the four (6.8k GitHub stars, against 1.7k for PegaProx and ProxCenter). ProxTrace is not yet commercially available.
- Covering other platforms: Pulse also follows Docker, Kubernetes and TrueNAS; PegaProx also manages XCP-ng; ProxCenter migrates from ESXi, XCP-ng, Nutanix and Hyper-V. ProxTrace only covers Proxmox VE.
What ProxTrace does differently
- The before and after value of every change, with the raw observation, linked to the machine, the account and the source address when the match is unique. None of the other three public pages describes this; PegaProx detects drift from a baseline, which says that a setting moved, not necessarily who changed it or when. See who changed this VM?
- Read-only as a principle: the agent cannot change anything, which makes its security review simpler (detailed rights).
- Missing data is never a success: a check that cannot be assessed stays “Not assessable”, and sealed reports keep a date’s state exactly as it was assessed (audit method).
When to use them together
- Pulse and ProxTrace: Pulse to know what is wrong right now, ProxTrace to know what changed before and prove it. Both read the infrastructure without changing it by default.
- PegaProx or ProxCenter and ProxTrace: the management tool to act, ProxTrace to keep an independent record of what was done, including by the management tool itself, directly in Proxmox VE or by a script.
- For an audit: the Proxmox VE hardening checklist and the Proxmox VE and NIS2 guide show which evidence each tool can provide.
Neighbouring tools
Other projects cover part of the topic: PVEViewer (inventory and health check, free Windows application), cv4pve-report (cluster export to Excel, HTML or JSON), the Proxmox-Hardening-Guide (hardening guide based on CIS Debian) and the CIS benchmark applied to Proxmox VE. They complement one of the tools above more than they replace it.
Updates
This comparison will be reviewed every quarter, with a new survey date. If a cell is wrong or outdated, write to us: we will correct it with the source.